Nitrogen stable isotope (15N, 14N) natural abundance has been much less used than carbon isotopes (13C, 12C) in plant physiology and ecology. Analytical problems, the lower fractional abundance of 15N than of 13C in the biosphere, the greater complexity of the N cycle relative to the C cycle, and smaller expressed discriminations in nature, are contributing factors. The major N pools, globally, have different isotope signatures: atmospheric N2 is 15N-depleted relative to organic N (including sedimentary N), a situation resulting from a greater expressed discrimination in the organic N to N2 (via denitrification) reaction than of diazotrophy during accumulation of the reduced N. Essentially all of the enzymes except nitrogenase which transform N compounds show discrimination against 15N, although for glutamine synthetase, and the amination of 2-oxoglutarate and pyruvate, this is only seen in terms of NH4+ rather than the true substrate, NH3. Discrimination is expressed in various N interconversions within plants, leading to substantial differences in δ15N (up to 12‰) among N compounds and macroscopic plant parts. N isotope fractionation during assimilation of exogenous combined N is often much lower than that expected from studies of isolated enzymes due to processes which show very little discrimination, such as limitation by transport through aqueous solution and membranes. Application of 15N/14N discrimination studies to plant ecology have concentrated largely on distinguishing diazotrophy from N supplied from combined N, based on the lower 15N/14N in diazotrophs due to the higher 15N/14N of combined N sources not being offset by fractionation during uptake. While potentially very useful, a number of pitfalls are discussed in its ecological use in both terrestrial and aquatic systems. N isotope discrimination is also useful in tracking N through food webs, and hence, back to combined N sources for plants.